This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: m68k ELF: e_flags bit for m68000 executables?
- From: Matt Fredette <fredette at theory dot lcs dot mit dot edu>
- To: nickc at cambridge dot redhat dot com (Nick Clifton)
- Cc: binutils at sources dot redhat dot com, fredette at netbsd dot org
- Date: Tue, 18 Dec 101 11:13:30 -0500 (EST)
- Subject: Re: m68k ELF: e_flags bit for m68000 executables?
> Hi Matt,
>
> > binutils' include/elf/m68k.h defines EF_CPU32, and I could use
> > an EF_M68000.
>
> Any particular reason why ?
I'm working on NetBSD for the sun2 machines, which only have 68010s.
The 68000 and 68010 are missing a small number of instructions and
addressing modes that appeared starting on the 020, so I need to know
if a binary is "safe" for me to run.
> > I don't have the m68k SYSV ABI supplement; does it specify such a
> > flag?
>
> Unknown. The ABI does not appear to be in the public domain, so I
> cannot answer this question.
>
> > If not, since I bet the m68k ABI is ownerless, can we
> > agree on a value?
>
> How about 0x00100000 ? [Chosen not to intersect with EF_CPU32 incase
> those two bits have extra significance not documented in the sources.
> Chosen to be higher than EF_CPU32 in case the lower bits have some
> significance in the (missing) ABI.]
Sounds good to me. I'll put together a patch shortly.
> > (Why is EF_CPU32 not a single bit?)
>
> I have no idea - it does seem very strange though.
Thanks!
Matt
--
Matt Fredette
http://mit.edu/fredette/www